The US stands out among developed economies for its comparatively low level of redistribution as a percentage of GDP. Gustavo de Souza examines the economic, institutional, and ideological determinants of voters’ preferences for redistribution and finds ideology, particularly views toward race and welfare recipients, significantly explains why voters do not desire higher levels of redistribution.
Redistributive policies, such as welfare transfers and universal healthcare, are often at the center of political debate. The US government distinguishes itself from other governments around the world in this category. While some EU countries spend as much as 35% of their GDP on transfers, the US has recently reached its historical high at only 20%. This has left many economists wondering why the US comparatively redistributes so much less.
Economists have come up with several potential explanations. One theory says that the median voter does not see economic gains from an increase in redistribution. For example, an increase in redistribution financed via higher taxes may lead to a significant drop in investments or employment, which would outweigh the benefits of increased redistribution to those in the lower- income brackets.
Another theory is that some voters, due to distrust of the government or discrimination, hold negative ideologies against government intervention and do not want to see an increase in redistribution. For instance, racial animosity may dissuade white voters from supporting an increase in redistribution that could benefit Black citizens.
A third theory says that the two-party system in the US might lead voters to focus on issues other than redistribution when choosing candidates, resulting in a level of redistribution that does not reflect the preferences of voters. For instance, if voters are too focused on issues such as abortion or gun control laws, they may choose a party based on their stance on these issues rather than their redistributive platform.
Lastly, some economists believe that institutional factors, such as low voter turnout among low-income voters and campaign contributions by the rich, bias redistribution in the US towards the preference of high-income voters.
It’s crucial to understand which of these theories best explains the political choice of redistribution in the US, as the answer has a direct effect on policy recommendations. If the main driver of redistribution is institutional factors, then a mandatory voting policy or cap on campaign contributions could be highly impactful. Alternatively, if voters don’t want redistribution for ideological reasons, understanding and addressing these reasons should be a priority. It could also be the case that the economic gains from redistribution are small, implying that no political reform is necessary.
To understand the key drivers of redistribution in the US and advise policy reform, I built and estimated a quantitative model bringing together features of macroeconomics, political economy, and industrial organization. In the model, agents vote to choose the level of redistribution and other policies to be implemented. They make their standard economic decisions, choosing labor supply, consumption, and savings, and engage in politics by choosing if they vote and which party to vote for. Parties choose their tax proposal to maximize their probability of winning the election and differ in other policy proposals.
The model, along with survey data, identifies the different channels affecting the political choice of redistribution in the US. It provides an estimate of the economic gains of redistribution, or, in other words, how a marginal increase in taxes would affect individuals at different income levels. It also provides a measure of the importance of ideological views on redistribution, and a measure of the preference of voters for other policies proposed by parties.
I found that voters could economically benefit from an increase in redistribution but they vote against it predominantly for ideological reasons. If agents were to vote disregarding their ideological views on redistribution, redistribution would increase from 17% of GDP to 36% of GDP. In contrast, if they voted disregarding their economic benefit from redistribution, the equilibrium redistribution as share of GDP would drop from 17% to 15%. If voters were to choose a level of redistribution that maximized their economic gain, it would increase to 35% of GDP.
Imposing compulsory voting or capping campaign contributions would not have any effect on the equilibrium government size. Because economic gain is not an important component of preferences for redistribution, the preferred tax rate does not differ significantly across the income distribution. What that means is that rich voters and poor voters would like almost the same level of redistribution. Therefore, reducing the influence of the rich by increasing voter turnout or capping campaign spending would not have a significant effect on the equilibrium tax rate.
I also found that racial animosity is one of the most important values driving voters’ ideological views on redistribution. Voters who hold negative views of Black and Hispanic Americans are more likely to oppose an increase in redistribution. If agents were to vote disregarding their racial views, redistribution would increase by 80%. In the same fashion, voters who hold negative views about welfare recipients are more likely to oppose an increase in redistribution, but of relatively smaller importance. If agents were to vote without negative views about welfare recipients, redistribution would increase by 27%.
Other party policies, despite not being as important as voters’ ideological views on redistribution, do have a sizable effect on redistribution. If it wasn’t for parties’ positions on gun control, abortion, and other issues unrelated to redistribution, redistribution in the US would increase from 16% to 19% of GDP, on par with the European average.
While views on race, and welfare recipients are important drivers of support for redistribution, more than 82% of the variance in preferences for redistribution cannot be explained by observable characteristics of voters. These results show that more research is needed to understand what drives people’s willingness to support redistribution.